JP6669738B2 - Surface inspection method for ceramics - Google Patents

Description

The present invention relates to a method for inspecting a surface of a ceramic body, and more particularly, to a method for detecting a crack formed on a surface of a ceramic body manufactured by using an extrusion molding method.

  2. Description of the Related Art Ceramic honeycomb structures are widely used as filters for collecting particulate matter contained in exhaust gas from internal combustion engines, boilers, and the like, and as catalyst carriers for exhaust gas purifying catalysts. The honeycomb structure is a tubular (eg, cylindrical) structure having both ends open and a so-called honeycomb structure (honeycomb structure) inside. That is, the honeycomb structure has a plurality of cells, each of which is partitioned by a partition wall, each extending along the axial direction of the structure, inside a cylindrical outer surface (outer wall). The honeycomb structure made of ceramics is widely used in the above-mentioned applications and the like because it is excellent in heat resistance, thermal shock resistance, and oxidation resistance.

  In general, a ceramic honeycomb structure is formed by molding a clay-like clay obtained by kneading ceramic (eg, alumina) powder as a constituent material thereof with an organic binder, water, or the like, by an extrusion molding method. It is manufactured by firing the honeycomb formed body thus obtained.

  However, when a honeycomb structure is manufactured by such a method, a defect such as a crack may be generated on an outer wall of the honeycomb structure or a foreign substance may adhere to the honeycomb structure. The generation of cracks and the attachment of foreign matter cause a decrease in the strength of the honeycomb structure, a decrease in filtration performance when the honeycomb structure is used as a filter, and a decrease in exhaust gas purification performance when used as a catalyst carrier. May be invited. Therefore, prior to using the honeycomb structure, it is necessary to inspect the presence or absence of those defects. Techniques for such a defect inspection are already known (for example, see Patent Documents 1 to 3).

  In particular, Patent Document 1 and Patent Document 2 image a surface of an outer wall which is a side surface of a honeycomb structure while rotating the cylindrical honeycomb structure around a central axis thereof, and detect defects on the outer wall based on the imaging result. It discloses a method for checking for the presence or absence of the presence.

  Since the presence of a defect as described above affects the characteristics of the honeycomb structure, the defect inspection is important in maintaining the quality of the honeycomb structure. In particular, when a cylindrical honeycomb structure is produced as a fired body of ceramics by firing an extruded honeycomb formed body, cracks along the axial direction of the honeycomb structure are easily formed, and such cracks are preferably formed. It is required to detect.

  On the other hand, in a cylindrical honeycomb structure which is a fired body of such a ceramic, when the outer wall is viewed along the circumferential direction, undulation (surface undulation) may be present. Such undulations are acceptable because they do not affect the quality and properties of the honeycomb structure.

  However, when a defect inspection is performed by the methods disclosed in Patent Literature 1 and Patent Literature 2, for example, even a wave that does not need to be determined as a defect is detected as a defect. ) Occurs.

  In particular, the shadow formed at the location of the crack extending in the axial direction of the honeycomb structure is similar to the shadow formed at the location of the concave portion formed in the axial direction of the honeycomb structure due to the undulation. Tends to be overdetected as the former.

  Inspection by visual inspection has a high degree of certainty in crack detection, but requires an inspection time, which is disadvantageous in terms of production efficiency and cost.

International Publication No. 2007/105825 US Patent Application Publication No. 2011/0128370 JP-A-09-257671

The present invention has been made in view of the above problems, and has as its object to provide a surface inspection method that can more reliably determine the presence or absence of cracks formed on the surface of a ceramic body than before. .

According to a first aspect of the present invention, there is provided a method of inspecting a crack on an outer wall surface which is a side surface of a cylindrical ceramic body, wherein a center axis of the table is rotatable in a horizontal plane, the center axis is in a vertical direction, and the rotation of the table is performed. A mounting step of mounting and holding the ceramic body so as to coincide with a center; and a first illumination light having a different wavelength band with respect to a predetermined irradiated area of a surface to be inspected as the outer wall surface of the ceramic body. In a state where the irradiation direction is simultaneously irradiated with the second illumination light in a mode in which the irradiation direction is included in one horizontal plane perpendicular to the central axis, the predetermined one of the imaging units arranged in the one horizontal plane An imaging step of imaging an irradiated area and generating one imaging data of a predetermined data format as one imaging result; and a determination image that can be used to determine the presence or absence of a crack based on the imaging result in the imaging step. To A determination image generation step to be performed, and a determination step of determining the presence or absence of a crack on the surface to be inspected based on the determination image, wherein the first and second illumination lights have respective optical axes. Irradiation is performed on the surface to be inspected from different directions sandwiching the imaging unit so that the angle between the center and the normal to the outer wall is the same angle θ in the range of 5 ° ≦ θ ≦ 30 °. In the imaging step, while the ceramic body is rotated once around the central axis by rotating the table, a plane parallel to the central axis is used as the surface to be inspected. The one image data is obtained by performing imaging by means, and the determination image generation step includes forming a pixel value of a predetermined color component from the one image data by forming an image of the color component. Signal and A determination image data generating step of generating determination image data that is image data of the determination image by obtaining the image data of the first color component. An image consisting of only a formation signal, or an image formation signal of the first color component and an image formation signal of a color component other than the first color component, the image having a signal amount equal to or less than a predetermined threshold value First determination image data that is image data of a first determination image that is an image composed of a formation signal, and image formation of a second color component having a wavelength range different from the first color component An image consisting of only signals, or an image forming signal for the second color component and an image forming signal for a color component other than the second color component, the signal being a predetermined threshold or less And second determination image data that is image data of a second determination image that is an image composed of the image formation signals of the first and second images. It is an image from which the presence or absence of a crack on the surface to be inspected can be determined based on the difference in the formation mode of the shadow area in each of the comparisons, and in the determination step, the first and second image data for determination By determining whether the shadow area is formed differently in each of the first and second determination images by comparing the first and second determination images , the presence or absence of a crack on the inspection surface is determined, and the result of the determination is determined. It is adapted to generate a determination result data describing the first and on the basis of the second determination image data, of the ceramic body in the first and the second determination image When it is determined that there is a shadow area having a smaller pixel value than its surroundings extending along the same direction at the same position on the inspection surface, and one of the first and second determination images On the other hand, when it is determined that the shadow region exists but the region corresponding to the shadow region in the other determination image does not exist at any of the formation position of the shadow region and its vicinity, It is determined that a crack along the same direction has occurred at a position corresponding to the shadow area on the inspection surface .

According to a second aspect of the present invention, in the method for inspecting a surface of a ceramic body according to the first aspect, in the imaging step, an imaging result by the imaging means is such that a pixel value for each of a plurality of color components is independent. are generated as image pickup data of the data format described Te, in the judgment image data generation process, from the previous SL iMAGING data by acquiring only the pixel values for the first color component, or the By acquiring a pixel value of a first color component and a pixel value of a color component other than the first color component and equal to or less than a predetermined threshold value, the first determination image data is obtained. generated from the previous SL IMAGING data by acquiring only the pixel values for the second color component, or the color components other than the pixel value and the second color component of the second color component Where the pixel value By obtaining the following pixel values threshold, to generate the second determination image data, and so on.

According to a third aspect of the present invention, in the method for inspecting a surface of a ceramic body according to the first aspect, in the imaging step, an imaging result by the imaging means is combined with pixel value information on a plurality of color components. It is generated as image pickup data of the data format, the in the determination image data generation process by breaking down the previous SL iMAGING data, either write only pixel values for the first color component, or the first Generating the first determination image data in which a pixel value of one color component and a pixel value of a color component other than the first color component and not more than a predetermined threshold value are described. , by decomposing before Symbol IMAGING data, the second either write only pixel values for the color components, or the color component other than the pixel value and the second color component of the second color component Pixel value And have values less than pixel values is described, to generate the second determination image data, and so on.

According to a fourth aspect of the present invention, in the surface inspection method for a ceramic body according to any one of the first to third aspects, the wavelength range of the first color component is equal to the wavelength band of the first illumination light. The wavelength range of the second color component at least overlaps with the wavelength band of the second illumination light.

According to a fifth aspect of the present invention, in the surface inspection method for a ceramic body according to any one of the first to fourth aspects, in the imaging step, the ceramics rotated around the central axis by the imaging means. By repeatedly imaging the body so that the imaging ranges are adjacent or partially overlapping at a predetermined imaging width, the entire surface to be inspected is imaged, and a plurality of images obtained by the repeated imaging by the imaging unit are obtained. The first and second imaging results are obtained by synthesizing the captured images.

According to a sixth aspect of the present invention, in the surface inspection method for a ceramic body according to the fifth aspect, a line sensor having sensitivity to at least the first and second illumination lights is used as the imaging unit. I did it.

According to a seventh aspect of the present invention, in the method for inspecting a surface of a ceramic body according to any one of the first to sixth aspects, the first and second determination images are displayed on a predetermined image display means by the crack. And an image display step of displaying the presence / absence of the image in a determinable manner.

According to an eighth aspect of the present invention, in the surface inspection method for a ceramic body according to the seventh aspect, in the determining step, the first and second determination images displayed on the image display means are compared. By doing so, the presence or absence of a crack on the surface to be inspected is determined by determining the difference in the formation of the shadow area in each of the first and second determination images.

According to a ninth aspect of the present invention, in the surface inspection method for a ceramic body according to the eighth aspect, in the determining step, the first and the second are determined based on the first and the second determination images. (2) when it is determined in the image for determination 2 that a shadow area darker than the periphery extending along the same direction exists at the same position on the surface to be inspected of the ceramic body, and the first and the second It is determined that the shadow region exists in one of the determination images, but the region corresponding to the shadow region in the other determination image does not exist in any of the formation position of the shadow region and its vicinity. In this case, it is determined that a crack along the same direction has occurred at a position corresponding to the shadow area on the surface to be inspected of the ceramic body.

A tenth aspect of the present invention is the method for inspecting a surface of a ceramic body according to any one of the first to ninth aspects, wherein the wavelength band of the first illumination light is 400 nm to 500 nm, and the second illumination The wavelength band of the light was set to be 600 nm to 800 nm.

According to an eleventh aspect of the present invention, in the surface inspection method for a ceramic body according to any one of the first to ninth aspects, the wavelength band of the first illumination light is 100 nm to 400 nm, and the second illumination The wavelength band of light was 300 nm to 800 nm.

According to a twelfth aspect of the present invention, in the method for inspecting the surface of a ceramic body according to any one of the first to eleventh aspects, there is provided a honeycomb structure obtained by firing a ceramic molded body obtained by extruding the ceramic body. The side surface of the honeycomb structure is the inspection surface.

According to a thirteenth aspect of the present invention, in the surface inspection method for a ceramic body according to any one of the first to twelfth aspects, the imaging step and the image generation step for determination are performed by the table and the holding unit. First illuminating means capable of irradiating the first illuminating light to the predetermined illuminated area of the ceramic body held in the ceramic body, and the predetermined illuminated area of the ceramic body held in the holding section A second illuminating unit capable of irradiating the second illuminating light, an imaging unit, and the first determination image data and the second determination image data based on the one imaging result. A first image and a second image illuminating means are arranged in the one horizontal plane with the imaging means interposed therebetween, and the first illuminating means and the second illuminating means are arranged with respect to the surface to be inspected from different directions. The first and the second Irradiating the illumination light is performed by the surface inspection apparatus, and so.

According to the first to thirteenth aspects of the present invention, an imaging method using two illumination lights having different irradiation directions and a simple method of comparing two determination images generated based on the imaging result, It is possible to more reliably determine the presence or absence of cracks on the surface to be inspected of the ceramic body than before.

Further , according to the first to thirteenth aspects, the presence or absence of cracks on the surface to be inspected of the ceramic body can be reliably determined by automatic processing.

Further , according to the first to thirteenth aspects, it is possible to reliably determine whether the deformation of the surface to be inspected of the ceramic body is a crack or a swell along a predetermined direction. Become. Thus, cracks are reliably detected, and undulations are not erroneously detected as cracks, thereby suppressing overdetection.

According to the first to thirteenth aspects, two images for determination only need to be taken once for the entire surface to be inspected by imaging using two illumination lights having different wavelength bands in addition to the irradiation direction. Are obtained as images at completely the same position, so that the presence or absence of a crack can be more accurately determined by a simple method of comparing two determination images.

Further, according to the first to thirteenth aspects, it is possible to more efficiently image the inspection surface of the cylindrical ceramic body.

Further, according to the first to thirteenth aspects, two images for determination can be obtained as images at completely the same position by simply rotating the cylindrical ceramic body, so that it is quicker and more accurate. It is possible to determine the presence or absence of a crack.

In particular, according to the seventh to ninth aspects, it is possible to confirm the presence or absence of cracks on the surface to be inspected of the ceramic body by a simple method of comparing the two determination images displayed on the image display means. Becomes

In particular, according to the twelfth aspect, it is possible to reliably determine whether the deformation that has occurred on the surface to be inspected of the ceramic body is a crack or undulation along the extrusion direction during extrusion molding. Becomes Thereby, cracks along the extrusion direction are reliably detected, and undulations are not erroneously detected as cracks, so that overdetection is suppressed.

FIG. 1 is a diagram showing a surface inspection apparatus 1 together with a honeycomb structure 100 to be inspected. FIG. 2 is a block diagram illustrating components included in the surface inspection apparatus 1. It is a figure which shows the type of the deformation which arises in the outer wall 101 of the honeycomb structure 100 along the axial direction by the enlarged view of the part F in the cross section perpendicular to an axial direction. It is a figure for showing how an image is formed in a portion of outer wall 101 where crack CR is formed. It is a figure showing an image of a place where crack CR is formed about actual outer wall 101. FIG. 4 is a diagram showing how an image is formed at a location on an outer wall where a undulation UD is formed. FIG. 9 is a diagram illustrating an image of a portion where an undulation UD is formed on an actual outer wall 101. FIG. 4 is a diagram illustrating how an image is formed at a location where a step ST is formed on an outer wall 101. FIG. 9 is a block diagram illustrating components included in a surface inspection device 1 according to a modification.

<Surface inspection device and honeycomb structure>
FIG. 1 is a diagram showing a surface inspection apparatus 1 according to an embodiment of the present invention, together with a honeycomb structure 100 to be inspected. FIG. 1A is a diagram schematically illustrating a spatial arrangement relationship between components of the surface inspection apparatus 1 and the honeycomb structure 100. FIG. 1B is a plan layout view of a main part of the surface inspection apparatus 1. FIG. 2 is a block diagram illustrating components included in the surface inspection apparatus 1.

The surface inspection device 1 is for inspecting the presence or absence of a defect occurring on the outer wall 101 using the surface (outer surface) of the outer wall 101 as a side surface of the cylindrical honeycomb structure 100 as a surface to be inspected. The surface inspection apparatus 1 roughly captures an image of the surface of the outer wall 101 parallel to the central axis AX1 under predetermined illumination light while rotating the honeycomb structure 100 around its central axis AX1, and based on the imaging result. Thus, the presence or absence of a defect on the outer wall 101 is determined. Here, the honeycomb structure 100 is a cylindrical structure having both ends open and a so-called honeycomb structure (honeycomb structure) inside. The honeycomb structure 100 includes a plurality of cells 103, each of which is defined by a partition wall 102 and extends along the central axis AX1 (axial direction) of the honeycomb structure 100 inside a cylindrical outer wall 101. It is what you have. In FIG. 1B, only one cell 103 is painted out for the sake of understanding, but actually all the cells 103 penetrate in the axial direction.

  For example, the thickness of the outer wall 101 is about 500 μm to 1 mm, the thickness of the partition 102 is about 50 μm to 300 μm, and the pitch of the partition 102 that defines the size of the cell 103 is about 0.5 mm to 2.0 mm. is there. The length (height h) in the axial direction is about 40 mm to 400 mm, and the radius (cross-sectional radius) in a cross section perpendicular to the axial direction is about 20 mm to 200 mm.

  Note that, in the present embodiment, an embodiment in which the honeycomb structure 100 includes cells 103 having a square shape in cross section and a uniform size is shown, but this is an example, and the cells 103 have a regular hexagonal shape in cross section. Or may be circular in cross section, or may have cells 103 of different sizes.

  The honeycomb structure 100 according to the present embodiment is a fired body of ceramics (for example, alumina), and is a clay-like clay obtained by kneading ceramic powder as a constituent material thereof with an organic binder, water, and the like. It is manufactured by forming soil by an extrusion molding method and firing a honeycomb formed body (ceramic formed body) obtained by the method.

  Examples of the defect generated on the outer wall 101 of the honeycomb structure 100 include a crack and adhesion of a foreign substance. The surface inspection apparatus 1 according to the present embodiment includes at least the outer wall of the defects. Cracks generated along the axial direction of the honeycomb structure 100 in 101 are to be detected. In the present embodiment, a crack that occurs in such a manner is particularly referred to as an axial crack. Although details will be described later, in the axial crack, a part of the outer wall 101 shifts in the radial direction along the axial direction, in addition to a general crack CR (see FIG. 3) that causes a gap in the outer wall 101. A step (step crack) ST (see FIG. 3) resulting from the above is also included.

  As shown in FIG. 1A, a surface inspection apparatus 1 includes a table 2 on which a honeycomb structure 100 is mounted at the time of inspection, and a pair of illumination units that irradiate the honeycomb structure 100 mounted on the table 2 with illumination light. Illuminating means 3 (first illuminating means 3a and second illuminating means 3b), imaging means 4 for imaging the surface of outer wall 101 of honeycomb structure 100 to generate image data (imaging data), and surface inspection apparatus 1 A control unit 5 that controls the overall operation, performs image processing based on the image data obtained by the imaging unit 4, and determines various processing units that determine the presence or absence of a defect based on the processing result; Mainly prepare.

  The table 2 has a mounting surface on which the honeycomb structure 100 can be mounted in a posture in which the central axis AX1 substantially coincides with the vertical direction. The table 2 also includes a rotation mechanism 2a (see FIG. 2) not shown in FIG. With the rotation mechanism 2a, the table 2 is rotatable in a horizontal plane as indicated by arrows AR1 and AR2 in FIG. For example, a turntable is used for the rotation mechanism 2a. The table 2 can be said to be a part that functions as a holding unit that holds the honeycomb structure 100 so as to be rotatable around its central axis AX1.

  The rotation mechanism 2a includes an encoder (not shown). Each time the table 2 rotates by a predetermined angle, the encoder 2 sends a pulse (encoder pulse) to the imaging unit 4 (more specifically, the imaging control unit 4c). Is given.

  The honeycomb structure 100 is mounted on the mounting surface of the table 2 such that the central axis AX1 coincides with the rotation center of the table 2. In addition, as the table 2, a table whose plane size is smaller than a cross-sectional size perpendicular to the central axis AX <b> 1 of the honeycomb structure 100 is used. This is to prevent the image of the table 2 from being reflected when the image is taken by the image pickup means 4.

Forming the illumination means 3 of a pair and a first illumination means 3a and the second illumination means 3b, as shown in FIG. 1 (b), the outer wall 101 of the honeycomb structure 100 placed on the table 2 in a plan view They are arranged at positions symmetrical with respect to the direction of a certain normal line N. The first illumination means 3a and the second illumination means 3b irradiate the former illumination light (first illumination light) La and the latter illumination light (second illumination light) Lb to the same position on the outer wall 101. In other words, in other words, in the planar view, each irradiation angle (the optical axis centers AX2a and AX2b of the illumination light and the normal N Are formed so as to be the same angle θ. By satisfying such an arrangement relationship, the first illumination unit 3a and the second illumination unit 3b are disposed with the imaging unit 4 interposed therebetween, and the first illumination unit 3a and the second illumination unit 3b are arranged in the first direction with respect to the side surface of the honeycomb structure 100 that is the inspection surface from different directions. The illumination light La and the second illumination light Lb can be irradiated. The angle θ is a value in a range of 5 ° ≦ θ ≦ 30 °. In the following, unless otherwise specified, the overlapping area of the respective illuminated areas of the first illumination light La and the second illumination light Lb is simply referred to as the illuminated area.

  However, as the first illuminating means 3a and the second illuminating means 3b, those having different (non-overlapping) wavelength bands of illumination light emitted from each other are used. Further, in the irradiated area, the first illumination light La and the second illumination light Lb having different wavelength bands are irradiated in a superimposed manner.

  For example, in a preferred embodiment, the first lighting means 3a emits red light as first lighting light La, and the second lighting means 3b emits blue light as second lighting light Lb. Alternatively, the first illumination unit 3a may emit white light as first illumination light La, and the second illumination unit 3b may emit ultraviolet light as second illumination light Lb. Note that, in this embodiment, red light is light having an emission wavelength in a wavelength band of 600 nm to 800 nm, and blue light is light having an emission wavelength in a wavelength band of 400 nm to 500 nm. . In addition, it is assumed that white light is light having an emission wavelength in a wavelength band of 300 nm to 800 nm, and ultraviolet light is light having an emission wavelength in a wavelength band of 100 nm to 400 nm.

  As a preferred example of the illumination light source provided in the illumination means 3, an LED that emits monochromatic light that satisfies the above-mentioned wavelength band is used. However, for example, a mode in which light emitted from an LED, a metal halide lamp, or another light source that emits white light passes through a color filter or the like that passes only light that satisfies the above-described wavelength band, and then is irradiated onto the honeycomb structure 100 It may be.

  If an LED is used as the illumination light source, the illumination unit 3 further includes a not-shown condenser lens for condensing light emitted from the LED, and the light passing through the condenser lens irradiates the honeycomb structure 100. This is preferable from the viewpoint of increasing the irradiation intensity of the illumination light.

The imaging unit 4 is arranged such that at least the light receiving unit 4a is disposed on a normal line N to the outer wall 101 of the honeycomb structure 100 which is a symmetric axis of the arrangement position of the first illumination unit 3a and the second illumination unit 3b. That is, at the arrangement position of the imaging means 4, an image of the region to be illuminated by the illumination light emitted from the illumination means 3 at the normal line N can be captured within the range of the imaging width w.

  Although not shown in FIG. 1, the imaging unit 4 includes an imaging unit 4b and an imaging control unit 4c in addition to the light receiving unit 4a, as shown in FIG. The imaging unit 4b is a part that performs an actual imaging operation (imaging based on an imaging instruction) in the imaging unit 4. The imaging control unit 4c combines the imaging instruction to the imaging unit 4b, the acquisition of the imaging data from the imaging unit 4b, and the acquired image data for each imaging width w to perform one imaging for the entire surface of the outer wall 101. This part is responsible for processing such as generating continuous image data as image data.

  In the surface inspection apparatus 1 according to the present embodiment, the outer wall 101 of the honeycomb structure 100 that is rotated in a horizontal plane by the table 2 under the illumination light from the pair of illumination units 3 using the imaging unit 4. By repeatedly performing imaging for each predetermined imaging width w on the surface, a captured image of the entire outer wall 101 is finally obtained. This is realized by rotating the table 2 at a predetermined rotation speed (angular speed) and repeatedly performing imaging at a fixed timing (time interval) in the imaging unit 4b. The plurality of captured images obtained for each of the captured widths w are subjected to image processing in various processing units provided in the surface inspection apparatus 1 as described later, and presence / absence of a defect (in particular, a crack) on the outer wall 101 of the honeycomb structure 100. Is used to determine

  Note that the imaging width w and the distance (imaging distance) between the imaging unit 4 and the honeycomb structure 100 may be appropriately determined according to the cross-sectional radius of the honeycomb structure 100. That is, if the imaging width w is too large as compared with the cross-sectional radius, the ratio of the focused range to the imaging width w in the captured image becomes small, which is not preferable. Also, if the imaging width w is too small or the imaging distance is too large, the resolution of the captured image deteriorates, which is not preferable. When the cross-sectional radius is about 20 mm to 250 mm as described above, the imaging width w is preferably about 10 μm to 3 cm, and the imaging distance is preferably about 30 cm to 200 cm.

  More specifically, the imaging unit 4 (imaging unit 4b) is good for each wavelength band of the first illumination light La emitted by the first illumination unit 3a and the second illumination light Lb emitted by the second illumination unit 3b. In addition to having sensitivity, an image forming signal for a color component corresponding to a predetermined wavelength range can be generated and transferred from the imaging control unit 4c to the control unit 5. Here, the image forming signal is a signal representing a coordinate position and a signal amount at the coordinate position of the captured image. This is realized, for example, by the imaging unit 4 generating imaging data in a data format that independently (individually) describes a pixel value (color density value at each pixel position) for each color component. In such a case, the pixel position of each pixel corresponds to the coordinate position in the image forming signal, and the pixel value corresponds to the signal amount at the coordinate position. Hereinafter, unless otherwise specified, the case where the imaging unit 4 generates imaging data in a data format in which a pixel value for each color component is independently described in this manner will be described.

  As such an imaging unit 4, a line sensor that has individual light receiving sensitivity to light of different color components (wavelength ranges) and that can output imaging data in which pixel values of each color component are described independently is provided. Is exemplified. For example, as described above, if the first illumination light La is red light and the second illumination light Lb is blue light, a known RGB line sensor can be used as the imaging unit 4. Note that the RGB line sensor can receive green light in addition to red light and blue light, and outputs RGB format data as imaging data. However, information on the light intensity value (pixel value) of green light is described in this embodiment. Not used in form. When the line sensor is used as the imaging unit 4, it is preferable that the light receiving unit 4 a is arranged so that the longitudinal direction is parallel to the axial direction of the honeycomb structure 100.

  However, the use of a line sensor as the imaging unit 4 is not essential, and an embodiment using an area camera having a two-dimensional (rectangular) imaging range, such as a digital camera, may be used.

  Further, as the image pickup means 4, image pickup data having a format in which information of pixel values of each color component is synthesized for each pixel position while individually having light receiving sensitivity to light of a different color component (wavelength range). May be used. However, in such a case, it is required that the pixel value information of each color component can be appropriately restored in the subsequent processing.

  Preferably, the imaging unit 4 is provided so as to be able to perform imaging with an imaging width w over the entire area in the axial direction of the honeycomb structure 100 (the range of the height h shown in FIG. 1). This can be realized, for example, by using a line sensor capable of imaging the imaging range as the imaging unit 4. In such a case, the illuminating unit 3 is also provided so that the illuminating light is irradiated at an irradiation intensity suitable for imaging by the imaging unit 4 at least over the entire area of the imaging range. When the illumination unit 3 and the imaging unit 4 satisfy these requirements, it is possible to image the entire surface of the outer wall 101 during one rotation of the honeycomb structure 100.

  However, such an aspect is not essential, and the imaging means 4 is provided so as to be relatively movable in the vertical direction with respect to the honeycomb structure 100 mounted on the table 2, and for each part obtained by dividing the range of the height h into several parts. Alternatively, the image may be taken at the same time. Further, the surface inspection apparatus 1 may include a plurality of imaging units 4 in the vertical direction, and each of the imaging units 4 may capture an image of a different range to capture an image of the range of the height h as a whole.

  The control unit 5 is configured by, for example, a computer including a CPU (not shown), a ROM, a RAM, an HDD, and other storage media. In the surface inspection apparatus 1, the operation and processing of each unit are enabled by the CPU executing an operation program stored in a storage medium in advance. Regarding the control means 5, if an electrical connection is secured between each of the above-described rotation mechanism 2a of the table 2, the pair of illumination means 3, and the imaging means 4, the position of the control means 5 is particularly limited. There are no restrictions.

  As shown in FIG. 2, the control unit 5 includes an input operation unit 6 including, for example, a keyboard and a touch panel, which enables the surface inspection apparatus 1 to perform various input operations such as an external inspection execution instruction and condition setting. And a display unit 7 composed of, for example, a liquid crystal display for displaying an operation menu and a test result.

  Further, the control unit 5 includes a general control unit 10, a rotation control unit 12, a lighting control unit 13, a determination image generation unit 20 as functional components realized by executing the operation program by the CPU. And a defect determination unit 24.

  The overall control unit 10 is a part that comprehensively controls the operation of each unit of the surface inspection device 1. That is, each unit of the surface inspection apparatus 1 performs its operation based on a control signal from the general control unit 10.

  The rotation control unit 12 is a part that controls the operation of the rotation mechanism 2a provided in the table 2 (the rotation of the table 2 and its stop). That is, in the surface inspection device 1, the rotation operation and the stop of the table 2 are controlled by the rotation instruction signal and the stop instruction signal given from the rotation control unit 12 to the rotation mechanism 2 a based on the control signal from the overall control unit 10. It is being realized.

  The illumination control unit 13 is a part that controls operations of the first illumination unit 3a and the second illumination unit 3b. That is, in the surface inspection device 1, an on / off instruction signal (lighting / lighting) of the illumination light given from the illumination control unit 13 to the first illumination unit 3a and the second illumination unit 3b based on the control signal from the overall control unit 10. By the light-off instruction signal), irradiation of the illumination light from the first illumination unit 3a and the second illumination unit 3b and its stop are realized.

  Based on the image data of the outer wall 101 of the honeycomb structure 100 obtained by the imaging unit 4, the image generation unit for determination 20 is an image data representing an image (image for determination) used for determining the presence or absence of a crack in the outer wall 101. It is a part responsible for the generation processing of (judgment image data). The determination image generation unit 20 mainly includes a decomposed image generation unit 22 and a filter processing unit 23.

  The decomposed image generation unit 22 is a part that acquires the image data generated by the imaging unit 4 and generates the first decomposed image data and the second decomposed image data from the image data. The first decomposed image data is intended to be generated mainly as image data of an image formed by irradiation of the first illumination light La, and the second decomposed image data is mainly generated by irradiation of the second illumination light Lb. It is intended to be generated as image data of an image formed by the above, but it is difficult for the defect determination unit 24 to determine the presence or absence of a crack by comparing two determination images in a manner described later. To the extent that this is not the case, it is permissible for each image data to include an image with another light such as external light.

  The filter processing unit 23 is a unit that performs an appropriate filtering process on two pieces of decomposed image data generated by the decomposed image generation unit 22 so as to obtain an image suitable for defect determination performed by the defect determination unit 24. Examples of such filter processing include binarization processing, shading correction, and contraction / expansion processing. The two pieces of decomposed image data that have been subjected to the filter processing are provided from the image generation unit for determination 20 to the defect determination unit 24 as image data for determination when determining the presence or absence of a crack.

  The defect determination unit 24 is a part that performs determination processing for determining the presence or absence of a defect on the outer wall 101. The defect determining unit 24 compares at least two determination images (more specifically, two determination image data) with respect to at least the presence or absence of an axial crack occurring on the outer wall 101, and forms a shadow region in each of the two determination images. A determination process is performed based on the difference between the modes. Preferably, when it is determined that an axial crack exists, the defect determining unit 24 specifies the type of the axial crack to be determined (whether the crack is a general crack CR or a step ST) as part of the determination process. ) And the location of the occurrence of the axial crack is also specified. The details of the determination process will be described later.

  However, the presence / absence of cracks occurring in other modes (for example, occurring in directions other than the axial direction), the attachment of foreign matter, and the like are also determined by the defect determination unit 24 in an appropriate processing mode. You may. The control means 5 may be provided with another functional component (not shown) necessary for that purpose.

<Deformation type of outer wall>
Next, prior to the description of the process of determining the presence or absence of an axial crack, a type of deformation that occurs in the outer wall 101 of the honeycomb structure 100 in the axial direction, which is a prerequisite for the determination process, will be described. FIG. 3 is a diagram illustrating the type of deformation that occurs on the outer wall 101 of the honeycomb structure 100 along the axial direction thereof by an enlarged view of a portion F in a cross section perpendicular to the axial direction.

  Ideally, it is desired that the honeycomb structure 100 be manufactured so that the outer wall 101 thereof has no defects and is uniformly smooth. However, in actuality, due to the manufacturing method, it is more specific. Due to the fact that the honeycomb structure 100 is manufactured by a process of molding by an extrusion molding method and subsequent firing, deformation may occur in the axial direction of the honeycomb structure 100 which is also the extrusion direction at the time of extrusion molding. Specifically, when a honeycomb formed body is obtained by extrusion molding, a problem that occurs along the axial direction of the formed body due to lack of moisture or the like, or the honeycomb formed body is fired to obtain a honeycomb structure 100. Later, when the temperature is decreased, shrinkage or the like generated in the honeycomb structure 100 causes such deformation.

  There are three specific types of deformation, shown as type (a), type (b), and type (c) in FIG.

First, type (a) is a general crack CR generated in such a manner that a gap is provided in the outer wall 101 along the axial direction. In most cases, the crack CR is formed to penetrate the outer wall 101 as shown in FIG. 3, but a gap that does not penetrate the outer wall 101 and is closed on the inner surface side also corresponds to type (a). It shall be. Most of the axial cracks generated in the honeycomb structure 100 are the type (a) cracks CR. Although two end faces relative to the outer wall 101 forming the cracks CR are shown as both flat in Figure 3, in fact, in an embodiment having a fine unevenness reflecting the grain shape of the ceramic In some cases, an end face is formed. Most of the axial cracks are formed as such cracks CR. The width of the crack CR along the circumferential direction is about 20 μm to 500 μm. On the other hand, the depth of the crack CR naturally becomes the thickness of the outer wall 101 when the crack CR penetrates, so that the thickness becomes the upper limit value, but the lower limit value when the crack CR does not penetrate is about 50 μm or more.

  Next, type (b) is a step (step crack) ST generated as a result of a part of the outer wall 101 shifting in the radial direction along the axial direction. Although FIG. 3 illustrates a case where the right side in the drawing is relatively high and the left side in the drawing is relatively low, naturally, the step ST may be formed in an opposite manner. Can occur. Although such a step ST is rarely generated as compared with the crack CR, it is the same as the crack CR in that a discontinuity is generated in the outer wall 101, and in the present embodiment, like the crack CR, Treat as an axial crack to be detected. The height of the step ST is about 50 μm to 1 mm.

  On the other hand, type (c) is a undulation UD that is an undulation that occurs along the circumferential direction on the surface of the outer wall 101. More specifically, FIG. 3 illustrates a concave portion formed so as to have a longitudinal direction in the axial direction of the honeycomb structure due to the occurrence of the undulation UD. Although the undulation UD may be formed as a convex portion, the following description is directed to the undulation UD in which the concave portion is formed along the axial direction for easy comparison with the above-described crack CR.

  The width of the undulation UD along the circumferential direction is about 200 μm to 1 mm. Further, the depth of the undulation UD is about 200 μm to 1 mm.

  In the honeycomb structure 100, the above three types of deformation occur. Among them, the surface inspection device 1 according to the present embodiment should surely detect the crack CR and the step ST, which are collectively referred to as axial cracks. The presence of the undulation UD is permissible because it does not affect the quality and characteristics of the honeycomb structure. In other words, it is required that the surface inspection apparatus 1 reliably detect an axial crack when determining a defect, but not perform overdetection that the undulation UD is an axial crack.

<Inspection process>
Next, a description will be given of a typical processing process for inspecting for the presence or absence of an axial crack, which is performed in the surface inspection apparatus 1 having the above-described configuration. In the following, for simplicity of description, unless otherwise specified, the first lighting means 3a emits red light as the first lighting light La, and the second lighting means 3b emits blue light as the second lighting light Lb. In both cases, the imaging unit 4 is an RGB line sensor that generates image data in RGB format, and the illuminating unit 3 and the imaging unit 4 are connected to the surface of the outer wall 101 which is a side surface of the honeycomb structure 100 during one rotation. It is assumed that imaging is performed for the whole.

  First, the honeycomb structure 100 to be inspected is placed and fixed on the table 2. The mounting of the honeycomb structure 100 on the table 2 may be performed manually by an operator, or may be in a mode in which a predetermined transport means provided outside the apparatus is automatically mounted.

  When such placement is performed, an inspection start instruction is given to the overall control unit 10 through the input operation unit 6. In addition, parameters necessary for the inspection, for example, the size of the honeycomb structure 100, the rotation speed of the table 2, the intensity of the illumination light, imaging timing (time interval) and other imaging conditions, filter processing conditions, and axial crack determination processing conditions Are set through the input operation unit 6 in advance. Note that a configuration in which the mounting and fixing of the honeycomb structure 100 is automatically detected and an inspection start instruction is given to the overall control unit 10 may be employed.

  Upon receiving the inspection start instruction, the overall control unit 10 controls the rotation control unit 12 and the illumination control unit 13 synchronously.

  Specifically, the general control unit 10 first instructs the rotation control unit 12 to rotate the rotation mechanism 2a provided on the table 2 on which the honeycomb structure 100 is placed, and also instructs the illumination control unit 13 to rotate. On the other hand, irradiation of the illumination light by the first lighting means 3a and the second lighting means 3b is instructed. The rotation control unit 12 and the illumination control unit 13 emit a drive signal in accordance with the instruction signal given from the general control unit 10, so that the table 2 on which the honeycomb structure 100 is placed has a predetermined rotation speed. At the same time, the first illumination light La and the second illumination light Lb having predetermined intensities from the first illumination means 3a and the second illumination means 3b are respectively transmitted to the rotating honeycomb structure 100. As shown in FIG. 1B, irradiation is performed superimposedly (simultaneously).

  When the rotation operation of the table 2 is started, an encoder (not shown) provided in the rotation mechanism 2a emits a pulse (encoder pulse) at a predetermined time interval. The encoder pulse is transferred to the imaging control unit 4c of the imaging unit 4. The imaging control unit 4c gives an imaging instruction to the imaging unit 4b so as to execute imaging in synchronization with the timing of receiving the encoder pulse. By performing imaging at the timing at which the encoder pulse is issued, the imaging unit 4 sets a data set of a pixel value for each RGB color component at the time of individual imaging and a pulse value of the encoder pulse at the time of imaging. , Imaging data is generated.

  Since the imaging unit 4 performs imaging with the imaging width w as described above, the cross-sectional radius of the honeycomb structure 100 is set to r, and the rotational angular velocity of the table 2 (hence, of the honeycomb structure 100) is set to ω (rad / Sec), at least 2πr / w times of imaging during a period of 4π / ω (second) during which the honeycomb structure 100 makes one rotation, ie, every 2w / rω (second) If one image is taken, it is possible to take an image of the entire surface of the outer wall 101 while the honeycomb structure 100 makes one rotation. Since it is preferable to have an overlap, it is preferable that the imaging time interval is set to a value smaller than 2 w / rω (second).

For example, when the honeycomb structure 100 with r = 5.0 cm is rotated at ω = 230 ° / sec and an image is taken by the imaging unit 4 (line sensor) with w = 30 μm, the imaging time interval is 1.0. It is set to about × 10 ⁻⁴ seconds to 1.4 × 10 ⁻⁴ seconds. In such a case, the number of times of imaging is about 10500 to 15000 times. That is, when inspecting one honeycomb structure 100, image data of about 10500 to 15000 lines is obtained.

The imaging data obtained by the imaging unit 4 is sequentially or collectively divided by a predetermined data amount (for example, each time the imaging of one honeycomb structure 100 is completed), and It is provided to the generation unit 22 .

  In addition, in the respective honeycomb structures 100, the determination of the completion of the imaging of the entire surface of the outer wall 101 is made by the overall control unit 10. More specifically, the overall control unit 10 determines the time required for imaging the entire surface of the outer wall 101 based on the size of the honeycomb structure 100, the rotation speed of the table 2, and the like, which have been input through the input operation unit 6 in advance before the inspection. Is calculated, and by acquiring the encoder pulse generated by the rotation mechanism 2a together with the imaging means 4, the timing at which the imaging is started can be known. The timing at which the imaging of the entire surface of the outer wall 101 is completed can be determined from the timing of starting the imaging and the time required for the imaging.

Alternatively, every time a certain number of encoder pulses corresponding to the rotation angle of the table 2 are input to the overall control unit 10 , the imaging data for a fixed number of times of imaging (the number of lines to be captured) is captured and set in advance. By automatically terminating the imaging based on the rotation angle of the table 2, the number of input lines, the number of encoder pulses, and the like, the outer wall 101 of the honeycomb structure 100 can be imaged at regular intervals.

  A large number of pieces of imaging data acquired from the start of imaging to the end of the imaging represent one piece of imaging data for the entire surface of the outer wall 101 by being arranged in chronological order. In the present embodiment, such one piece of imaging data is referred to as continuous image data.

  The integrated control unit 10 instructs the decomposed image generation unit 22 to generate decomposed image data based on the continuous image data at the timing when the imaging of the entire surface of the outer wall 101 is completed. Further, the overall control unit 10 gives an instruction signal to stop the rotation of the rotation mechanism 2a to the rotation control unit 12, and ends the irradiation of the first illumination light La and the second illumination light Lb to the illumination control unit 13. An instruction signal to be performed is given. The rotation control unit 12 and the illumination control unit 13 respectively emit drive signals in response to these instruction signals, so that the rotation of the table 2 is stopped, and the first illumination light La and the second illumination light Lb are turned off. You. Note that the first illumination light La and the second illumination light Lb may be in a constantly lit state. When the rotation of the table 2 is completely stopped, the honeycomb structure 100 for which imaging has been completed is moved from the table 2, and the honeycomb structure 100 to be imaged next is placed on the table 2 instead.

The decomposed image generation unit 22 generates first decomposed image data and second decomposed image data from the continuous image data generated as one piece of RGB image data. Now, since the first illumination light La is red light and the second illumination light Lb is blue light, the first decomposed image data is converted into image data (R image It is a preferable example to generate image data (B image data) by extracting only the B component from the continuous image data and generating the second decomposed image data as the second separated image data. For example, if the pixel value (color density value) at an arbitrary pixel (x, y) of the continuous image data is represented as (R _xy , G _xy , B _xy ), the decomposed image generation unit 22 R image data whose pixel value at (x, y) is expressed as (R _xy , 0, 0) and B image data whose pixel value at pixel (x, y) is expressed as (0, 0, B _xy ) Is preferably generated as first decomposed image data and second decomposed image data, respectively.

  In such a case, the R image data and the B image data are respectively image data obtained by irradiating only the first illumination light La and performing imaging by the imaging unit 4, and image data obtained by irradiating only the second illumination light Lb. 4 substantially corresponds to the image data obtained by performing the image pickup by the image pickup device 4.

However, as described above, if the pixel value is in a range that does not hinder the determination by the defect determination unit 24, for example, if the pixel value is equal to or less than a predetermined threshold value, the first decomposed image data is a color other than the R component. The second separated image data may include a color component other than the B component. That is, if the pixel value (color density value) at an arbitrary pixel (x, y) of the continuous image data is expressed as (R _xy , G _xy , B _xy ) as in the above-described example, The pixel value of the pixel (x, y) of the first decomposed image data is represented as (R _xy , g1 _xy , b1 _xy ), and the pixel value of the pixel (x, y) of the second decomposed image data is (r2 _xy , g2). _xy , _Bxy ). Here, g1 _xy , b1 _xy , r2 _xy , and g2 _xy each represent a pixel value equal to or less than a predetermined threshold value.

  Note that, as described above, the imaging unit 4 individually has light receiving sensitivity to light of different color components (wavelength ranges), and synthesizes pixel value information for each color component for each pixel position. In the case of using a device that outputs image data in a format, the decomposed image generation unit 22 restores pixel value information for each color component from the image data to obtain the first decomposed image data and the second decomposed image data. Generate

  The first decomposed image data and the second decomposed image data generated by the decomposed image generation unit 22 are subjected to a filtering process by a filter processing unit 23. The filter processing unit 23 performs the above-described binarization processing, shading correction, contraction / expansion processing, and the like on each of the first decomposed image data and the second decomposed image data.

  Then, the first decomposed image data and the second decomposed image data that have been subjected to the filter processing by the filter processing unit 23 are passed to the defect determination unit 24 as first determination image data and second determination image data, respectively. This is used for determining whether or not there is an axial crack in the defect determining unit 24. Details of the determination process in the defect determination unit 24 will be described later.

  The determination result data describing the determination result obtained by the defect determination unit 24 is transferred to the overall control unit 10 and further used for displaying the determination result on the display unit 7 and the like.

  When the determination result is obtained, subsequently, the inspection for another honeycomb structure 100 is similarly performed.

<Determination of axial crack>
The determination by the defect determination unit 24 is performed by roughly checking the images (the first determination image and the second determination image) represented by the first determination image data and the second determination image data, and appears in each image. This is performed based on the presence / absence of a shadow area formed due to the deformation occurring on the surface of the outer wall 101 and the difference in the formation position.

  First, where no deformation occurs, no shadow area is formed in each of the first determination image and the second determination image.

  Next, a case where a crack CR shown as type (a) in FIG. 3 has occurred as a modification will be described. FIG. 4 is a diagram showing how an image is formed at a portion of the outer wall 101 where the crack CR is formed.

FIG. 4A shows how an image is formed by the first illumination light La at a position where the crack CR is formed. When the first illumination light La is incident on the surface of the outer wall 101 at the predetermined angle θ as described above, in the areas RE11 and RE12 where the crack CR is not formed, the images of the surfaces by the first illumination light La are shown in the drawing. It is formed in the image pickup means 4 (not shown) arranged above the viewing position. In FIG. 4A, the image formation is indicated by solid arrows AR11 and AR12 (the same applies to FIGS. 6 and 8).

On the other hand, the region RE13 in which the crack CR is formed does not contribute to the image formation by the first illumination light La in the imaging unit 4. That is, the imaging unit 4 does not obtain an image of the region RE13, and the region becomes a shadow region. In FIG. 4A, the formation of such a shadow area is indicated by a dashed arrow AR13 (the same applies to FIGS. 6 and 8).

  On the other hand, FIG. 4B shows how an image is formed by the second illumination light Lb at a location where the crack CR is formed (the same location as in FIG. 4A). When the second illumination light Lb is incident on the surface of the outer wall 101 at the predetermined angle θ as described above, the regions RE21 and RE22 where no crack CR is formed (the same as the regions RE11 and RE12 in FIG. 4A, respectively) Regarding (2), as shown by arrows AR21 and AR22, an image of the surface by the second illumination light Lb is formed in the imaging unit 4.

  On the other hand, the region RE23 where the crack CR is formed (same as the region RE13 in FIG. 4A) does not contribute to the image formation by the second illumination light Lb in the imaging unit 4. That is, the imaging unit 4 does not obtain an image of the region RE23, and the region becomes a shadow region as indicated by the arrow AR23.

  FIG. 5 is a diagram illustrating an image of a portion where a crack CR is formed on the actual outer wall 101. FIG. 5A shows a captured image indicated by the captured data obtained by the imaging unit 4, and FIG. 5B shows a determination image generation unit based on the captured data giving the captured image of FIG. Reference numeral 20 denotes a first determination image based on the generated first determination image data, and FIG. 5C illustrates a second determination image based on the second determination image data similarly generated. That is, the ranges indicated by the three images are the same. Each image shown in FIG. 5 is monochrome for the sake of illustration, but is actually a color image (the same applies to FIG. 7).

Linear shadow regions S1 extending in the drawing viewing the vertical direction corresponds to the crack CR in FIG. 5 (a). On the other hand, the shadow area S1a formed in FIG. 5B and the shadow area S1b formed in FIG. 5C are respectively indicated by arrows AR13 and AR23 in FIGS. 4A and 4B. It corresponds to the represented shadow area. Comparing FIG. 5B and FIG. 5C, it can be seen that the shadow region S1a and the shadow region S1b have the same shape and formation position. For confirmation, the shapes and formation positions of these two regions also match the shadow region S1 in FIG.

  Next, a case where a undulation UD (recess) shown as type (c) in FIG. 3 is generated as a modification will be described. FIG. 6 is a diagram for illustrating how an image is formed at a portion of the outer wall 101 where the undulation UD is formed.

  FIG. 6A illustrates how an image is formed by the first illumination light La at a location where the undulation UD is formed. When the first illumination light La is incident on the surface of the outer wall 101 at the predetermined angle θ as described above, the first illumination light La does not constitute the undulation UD or faces the irradiation direction of the first illumination light La although it constitutes the undulation UD. As for the regions RE31 and RE32, as shown by the arrows AR31 and AR32, the image of the relevant surface is formed by the first illumination light La in the imaging unit 4.

  On the other hand, the region RE33 that forms the undulation UD and does not face the irradiation direction of the first illumination light La does not contribute to the image formation by the first illumination light La in the imaging unit 4. That is, the imaging unit 4 does not obtain an image of the region RE33, and the region becomes a shadow region as indicated by the arrow AR33.

  On the other hand, FIG. 6B shows how an image is formed by the second illumination light Lb at a location where the undulation UD is formed (the same location as in FIG. 6A). When the second illumination light Lb is incident on the surface of the outer wall 101 at the predetermined angle θ as described above, it does not constitute the undulation UD or faces the irradiation direction of the second illumination light Lb although it constitutes the undulation UD. In the regions RE41 and RE42, as shown by the arrows AR41 and AR42, an image of the relevant surface is formed by the second illumination light Lb in the imaging unit 4.

  On the other hand, the region RE43 constituting the undulation UD and not facing the irradiation direction of the second illumination light Lb does not contribute to the image formation by the imaging means 4 using the second illumination light Lb. That is, the imaging unit 4 does not obtain an image of the region RE43, and the region becomes a shadow region as indicated by the arrow AR43.

  FIG. 7 is a diagram showing an image of a portion of the actual outer wall 101 where the undulation UD is formed. FIG. 7A is a captured image indicated by the captured data obtained by the imaging unit 4, and FIG. 7B is a determination image generation unit based on the captured data giving the captured image of FIG. Reference numeral 20 denotes an image based on the first determination image data (first determination image), and FIG. 7C illustrates an image based on the second determination image data similarly generated (second determination image). is there. That is, the ranges indicated by the three images are the same.

Linear shadow region S2 that extends in the drawing viewing the vertical direction corresponds to the waviness UD in Fig 7 (a). On the other hand, the shadow area S2a formed in FIG. 7B and the shadow area S2b formed in FIG. 7C are respectively indicated by arrows AR33 and AR43 in FIGS. 6A and 6B. It corresponds to the represented shadow area. Comparing FIG. 7B and FIG. 7C, it can be seen that the formation positions of the shadow areas S2a and S2b do not match each other. For confirmation, the formation positions of these two regions are symmetric with respect to the shadow region S2 in FIG. 7A.

  As described above, the shadow region corresponding to the crack CR or the undulation UD is formed in both the first determination image and the second determination image at the location where the crack CR is formed and the location where the undulation UD is formed. However, in the case of the crack CR and the case of the undulation UD, the relationship between the formation positions of the shadow areas formed in the two determination images is different. That is, if the shadow area is for the crack CR, the shadow area in the first determination image matches the shadow area in the second determination image, but if the shadow area is for the undulation UD. The shadow area in the first determination image does not match the shadow area in the second determination image.

  Note that such a difference is obtained by utilizing a difference in the formation mode (formation size) of the crack CR and the undulation UD. Generally speaking, the crack CR tends to be formed thinner and deeper than the undulation UD.

  Further, a case where a step ST shown as type (b) in FIG. 3 occurs as a modification will be described. FIG. 8 is a diagram showing how an image is formed at a location on the outer wall 101 where the step ST is formed. In FIG. 8, it is assumed that the first illumination light La is emitted from the lower side of the step ST, and the second illumination light Lb is emitted from the upper side of the step ST.

FIG. 8A shows how an image is formed by the first illumination light La at a position where the step ST is formed. When the first illumination light La is incident on the surface of the outer wall 101 from the lower side of the step ST at the predetermined angle θ as described above, all the illuminated areas facing upward in the drawing in which the imaging unit 4 is arranged are arranged. In RE51 and RE52, as shown by arrows AR51 and AR52, an image of the surface by the first illumination light La is formed in the imaging means 4. Since there is no portion on the surface of the outer wall 101 where the first illumination light La is not irradiated, no shadow region is formed.

  On the other hand, FIG. 8B shows how an image is formed by the second illumination light Lb at a location where the step ST is formed (the same location as in FIG. 8A). As described above, when the second illumination light Lb is incident on the surface of the outer wall 101 from the upper side of the step ST at the predetermined angle θ, as described above, a region RE61 that is on the lower side of the step ST and is distant from the upper side, and the step ST As shown by arrows AR61 and AR62, an image of the relevant surface is formed by the first illumination light La in the imaging means 4 in the region RE62 which is the irradiation region on the upper side.

  On the other hand, on the lower side of the step ST, an area RE63 in which the second illumination light Lb is not irradiated is formed near the upper side, so that the imaging unit 4 does not obtain the image of the area RE63, As shown, the area is a shadow area.

  As a result, although illustration is omitted, when the step ST is formed in the mode shown in FIG. 8, the shadow area is not confirmed in the first determination image, and the shadow area of FIG. A shadow region similar to the region S2b is formed. The relationship between the shadow area in the first determination image and the formation position of the shadow area in the second determination image at the step ST is determined by the relationship between the shadow area in the first determination image and the second determination image for the crack CR and undulation UD described above. This is different from any relationship between the formation positions of the shadow areas in the image.

The angle range of the irradiation angle θ of the illumination light described above is determined from the viewpoint that a shadow region for discriminating the crack CR, the undulation UD, and the step ST is preferably formed. Specifically, when θ <5 °, the light source positions of the two irradiation lights are close to each other, and as a result, a shadow region of the step ST cannot be clearly obtained, which is not preferable. On the other hand, theta> If it is 30 °, is not preferable because the overlaps and the shadow region and the shadow region definitive the second determination image images in the first judgment image for waviness UD.

  The defect determination unit 24 determines the relationship between the formation position of the shadow region in the first determination image and the formation position of the shadow region in the second determination image as described above when the first determination image and the second determination image are collated. Based on the difference, the type of deformation occurring on the outer wall 101 is determined to be a crack CR, a step ST, or a undulation UD.

  In other words, the first determination image represented by the first determination image data and the second determination image represented by the second determination image data correspond to the same position of the outer wall 101 and correspond to the extrusion direction at the time of extrusion molding. When it is determined that there is a shadow area having a smaller pixel value than its surroundings extending along the direction in which the shadow area extends (usually the vertical direction in the first determination image and the second determination image), It is determined that a crack CR has occurred at the corresponding location.

  Further, although a shadow region exists in one of the first determination image and the second determination image, a region corresponding to the shadow region in the other image is located at any of the formation position of the shadow region and its vicinity. If not present, it is determined that a step ST has occurred at a location on the outer wall 101 corresponding to the shadow area. In such a case, it is also possible to specify the vertical relationship of the steps depending on which image has the shadow region formed.

  On the other hand, when a similar shadow area exists in both the first determination image and the second determination image, but the formation positions thereof are shifted, it is determined that the undulation UD has occurred on the outer wall 101.

  As described above, in the determination by the defect determination unit 24, the criterion for determining the crack CR and the step ST, which are the axial cracks, and the criterion for determining the undulation UD are clearly distinguished. Therefore, while the axial crack is reliably detected, the undulation UD is not erroneously detected as the axial crack. That is, in the determination by the defect determination unit 24, occurrence of overdetection is suitably suppressed.

Specifically, the defect determination unit 24 targets each of the first determination image data and the second determination image data generated by the determination image generation unit 20 by using a known image processing method. In the image represented by the data (the first determination image or the second determination image), a pixel range to be a shadow region is specified. Schematically, an area in which pixels having pixel values of 0 or a value within a predetermined threshold range close to 0 is continuous is specified as a pixel area corresponding to the shadow area. Then, the first determination image data and the second determination image data are collated, and based on the positional relationship between the pixel regions corresponding to the shadow region in each of the first determination image data and the second determination image data, the respective shadow regions have a crack CR, a step ST, and a undulation. It is determined which of the UDs corresponds. Then, if it exists shadow area is determined to correspond to the former two as axial crack is present, the defect determination unit 24, the type of pixel position information and axial crack (or cracks CR, The step ST) is described, and the determination result data describing the step ST) is generated and transferred to the overall control unit 10. The honeycomb structure 100 for which the determination result of the description content is created is determined to be a rejected product.

  When the pixel area corresponding to the shadow area is not detected in both the first determination image data and the second determination image data, and when it is determined that only the pixel area corresponding to the undulation UD exists, the axial crack is generated. Is not detected, the defect determination unit 24 generates determination result data describing that the honeycomb structure 100 to be inspected is a passed product, and transfers the data to the overall control unit 10.

  As described above, according to the surface inspection apparatus according to the present embodiment, the contrast between imaging using two illumination lights having different irradiation directions and two determination images generated based on the imaging results is compared. With this simple method, it is possible to more reliably determine the presence or absence of cracks on the side surface (outer wall surface) of the honeycomb structure than ever before. More specifically, the side surface (outer wall surface) of the honeycomb structure is imaged under two illumination lights belonging to two different wavelength bands, and subsequently, the obtained image data is used for each of the two wavelength bands. Whether the deformation generated on the outer wall of the honeycomb structure is a crack along the axial direction or a swell depending on how the shadow area is formed in the obtained two pieces of image data. Can be reliably determined. Thus, cracks are reliably detected, and undulations are not erroneously detected as cracks, thereby suppressing overdetection.

<Modification>
In the inspection method according to the above-described embodiment, the cylindrical honeycomb structure 100 is to be inspected. However, in principle of the inspection method, the honeycomb structure to be inspected is not necessarily cylindrical. . For example, the surface inspection apparatus 1 is provided with an advance / retreat moving mechanism for moving the image pickup means 4 with respect to the honeycomb structure placed on the table 2 so that the focal distance of the image pickup means 4 can be maintained. If the relative position between the imaging means and the honeycomb structure can be adjusted according to the imaging position, it is possible to inspect a honeycomb structure having a polygonal column shape, such as a rectangular cylindrical shape.

  In addition, on the principle of the inspection method according to the above-described embodiment, the inspection target does not necessarily need to be a honeycomb structure, but may be a cylindrical or columnar ceramic body that can be placed on a table. However, in view of the characteristic of the inspection method that it is possible to appropriately discriminate between cracks and undulations that occur in the axial direction, it is preferable that the ceramic fired body manufactured by applying the extrusion molding method is targeted.

  Alternatively, the surface inspection device 1 is provided with a parallel movement mechanism for moving the imaging unit 4 in parallel with the ceramic body placed on the table 2, and the imaging position is set so that the focal length of the imaging unit 4 is maintained. If the relative position between the imaging means and the ceramic body can be adjusted according to the above, inspection can be performed on a flat ceramic body.

  In the above-described embodiment, the first lighting means 3a and the second lighting means 3b forming a pair of lighting means 3 are arranged at positions symmetrical with respect to the direction of the normal line N of the outer wall 101 of the honeycomb structure 100. Thus, the respective irradiation angles are set to be the same angle θ. However, as long as the determination of the presence or absence of a crack according to the present embodiment can be appropriately performed, the first lighting unit 3a and the second lighting unit 3a can be used. The arrangement position of 3b may be asymmetric with respect to the direction of the normal line N. That is, the value of the irradiation angle θ with respect to the first lighting unit 3a and the second lighting unit 3b may be different as long as 5 ° ≦ θ ≦ 30 ° is satisfied.

  Alternatively, in the above-described embodiment, the imaging unit 4 is arranged on the normal line N. However, the imaging unit 4 does not necessarily have to be strictly provided on the normal line N. The imaging means 4 may be arranged at a position substantially coincident with the normal line or at a position near the normal line N as long as the determination of presence / absence can be suitably performed.

Further, in the above-described embodiment, the case where the R image data and the B image data are generated as the first decomposed image data and the second decomposed image data from one continuous image data is illustrated. By irradiating only the first illumination light La and irradiating only the second illumination light Lb and performing imaging by the imaging means 4 in each case, the R image data and the B image data are directly separated from the beginning as separate image data. It is also possible to perform the determination processing by the defect determination unit 24 using the generated information. However, in this case, after the honeycomb structure 100 is rotated by one rotation around the central axis AX1 , the image is captured by the imaging unit 4 while irradiating only the first illumination light La to obtain R image data. The B image data is obtained by taking an image with the imaging unit 4 while irradiating only the second illumination light Lb while rotating 100 further around the central axis AX1 . That is, it is necessary to rotate the honeycomb structure 100 at least twice. In addition, in that case, it is necessary to accurately position both images. In such a case, if the honeycomb structure 100 is displaced during rotation of the honeycomb structure 100, the rotation speed varies, or the eccentricity of the rotating shaft occurs, it becomes difficult to perform the alignment with high accuracy. Therefore, from the viewpoint of processing efficiency and judgment accuracy, to complete the imaging of only one rotation, and an R image data and B image data is data for exactly the same position generated based on one of the imaging data It is determined that the processing mode used in the above-described embodiment is superior.

  Furthermore, assuming that the wavelength bands of the two illumination lights are the same, the irradiation with only the first illumination light La and the irradiation with only the second illumination light Lb are performed. It is also possible to generate two image data corresponding to the B image data, collate them, and perform the determination processing by the defect determination unit 24.

  Further, in the above-described embodiment, red light and blue light are used as illumination light, and R image data and B image data are generated as decomposed image data. The wavelength band of the illumination light and the color component (the wavelength of Range) does not have to match. For example, while ultraviolet (UV) light and white light are used as illumination light, UV image data and R image data are generated as decomposed image data, and a determination image (determination image data) is generated based on the decomposed image data. May be generated.

Alternatively, as in the case described above, the honeycomb structure 100 is rotated twice around the central axis AX1 to perform imaging while irradiating only the first illumination light La in the first rotation, and the second illumination in the second rotation. In performing imaging while irradiating only the light Lb, each imaging result is generated as imaging data (first and second imaging data) in a data format in which a brightness value or a luminance value of each pixel is described, and determination is performed. The first image generation unit 20 generates the first and second determination images (first and second determination image data) by acquiring the brightness value or the luminance value of each pixel from the image data. Then, based on the first and second determination images (first and second determination image data), the presence or absence of cracks on the outer wall 101 of the honeycomb structure is confirmed. It may be.

  Alternatively, in the above-described embodiment, two illumination lights having different wavelength bands are used, and the image used for the determination is in accordance with those bands. However, two illumination lights having different polarization states are used instead. After performing imaging using each of the lights and imaging using each of two illumination lights having different phases, the deformation generated on the outer wall 101 due to the difference between the images formed by the two obtained images. May be a mode in which the type difference is specified.

  Further, in the above-described embodiment, the determination result obtained by the defect determination unit 24 is displayed on the display unit 7. However, instead of performing the determination by the defect determination unit 24, the determination image generation unit 20 are displayed on the display unit 7 based on each of the two determination image data generated by the display unit 20. The worker visually checks the two determination images to check the honeycomb structure. A mode in which the presence or absence of a crack in the outer wall 101 may be confirmed. If the two determination images are displayed on the display unit 7 so as to be able to compare the difference between the shadow regions existing in each of the two determination images, the presence or absence of a crack can be determined by a simple method of visual comparison.

  Further, in the above-described embodiment, mainly, the imaging result in the imaging unit 4 is described in the imaging data as a pixel value, passed to the determination image generation unit 20, and the determination image data is generated. The presence or absence of cracks in the outer wall 101 of the honeycomb structure is confirmed in a simple image processing mode. Instead, the imaging unit 4 converts the imaging result into, for example, analog image formation in a predetermined output format. The first determination image and the second determination image are displayed on the display unit 7 in a comparable manner using two different image forming signals of the image forming signals. May be adopted. For example, a case where an image signal in RGB format can be output by the imaging unit 4 and a first determination image including only the R signal and a second determination image including only the B signal are generated is exemplified. You. In such a case as well, the worker can visually check the two determination images to check for cracks in the outer wall 101 of the honeycomb structure.

  Alternatively, the determination of the presence / absence of a crack by the defect determination unit 24 and the determination of the presence / absence of a crack by the operator visually checking the two determination images displayed on the display unit 7 are used together in one surface inspection apparatus 1. It may be possible.

  Further, in the above-described embodiment, the imaging unit 4 performs imaging at a timing at which an encoder pulse is issued from the rotation mechanism 2a that rotates the table 2, but the imaging timing of the imaging unit 4 is controlled. The embodiment is not limited to this. FIG. 9 is a block diagram of the surface inspection apparatus 1 that controls the timing of imaging by the imaging unit 4 in a mode different from the above-described embodiment.

  Specifically, in the surface inspection apparatus 1 shown in FIG. 9, the operation of the imaging unit 4 is controlled by the imaging processing unit 14 provided in the control unit 5. In the surface inspection apparatus 1 described above, the imaging by the imaging unit 4 is performed by an imaging instruction signal given from the imaging processing unit 14 to the imaging unit 4 based on a control signal from the overall control unit 10.

  Further, in the surface inspection apparatus 1 shown in FIG. 9, the image generation unit for determination 20 includes the continuous image generation unit 21. The continuous image generation unit 21 is a part that performs processing of synthesizing data of images captured for each imaging width w to generate continuous image data that is one captured image data of the entire surface of the outer wall 101.

  When performing the inspection in the surface inspection apparatus 1, when the rotation operation of the table 2 and the irradiation state of the illumination light from the first illumination unit 3 a and the second illumination unit 3 b are stabilized, the general control unit 10 performs imaging. The processing unit 14 is instructed to execute imaging by the imaging unit 4. The imaging processing unit 14 causes the imaging unit 4 to perform imaging at a preset timing (time interval) in response to the execution instruction.

  When the imaging of the entire surface of the outer wall 101 is completed, the imaging processing unit 14 gives a signal to notify the general control unit 10 of the end. Upon receiving the notification signal, the general control unit 10 gives the rotation control unit 12 an instruction signal to stop the rotation of the rotation mechanism 2a, and also gives the illumination control unit 13 the first illumination light La and the second illumination light Lb. Is given. The rotation control unit 12 and the illumination control unit 13 respectively emit drive signals in response to these instruction signals, so that the rotation of the table 2 is stopped, and the first illumination light La and the second illumination light Lb are turned off. You.

  On the other hand, the imaging data obtained by the imaging unit 4 is sequentially or collectively provided to the continuous image generation unit 21 of the determination image generation unit 20 through the imaging processing unit 14.

  The continuous image generation unit 21 synthesizes the image data obtained at successive image capturing timings sequentially each time image data is obtained, or collectively after obtaining all image data, and outputs the continuous image data. Generate The generated continuous image data is provided to the decomposed image generation unit 22. Subsequent processing is performed in the same manner as in the above-described embodiment.

  Alternatively, the imaging control unit 4c itself may include a timer (not shown), and the imaging may be performed at a fixed timing using the timer.

However, the imaging method according to the above-described embodiment, in which imaging is performed every time the honeycomb structure 100 rotates by a certain angle , ensures that the outer wall 101 is securely positioned even when the honeycomb structure 100 is displaced during rotation. It can be said that the method is superior to the method of controlling the imaging timing according to these modifications in that the imaging can be performed. Further, in the above-described embodiment, the entire outer wall 101 of the honeycomb structure 100 is imaged by the single imaging unit 4, but instead, the periphery of the honeycomb structure 100 mounted on the table 2 is changed. Are mounted at predetermined intervals, and a first illumination means 3a and a second illumination means 3b are provided corresponding to each of the imaging means 4, and an imaging range of the plurality of imaging means 4 is provided. May cover the entire outer wall 101. In such a case, it is possible to obtain image data of the entire outer wall 101 of the honeycomb structure 100 without rotating the table 2.

Claims (13)

A method for inspecting a crack on an outer wall surface which is a side surface of a cylindrical ceramic body,
A mounting step of mounting and holding the ceramic body on a mounting surface of a table that is rotatable in a horizontal plane, such that a central axis thereof is vertically aligned with the rotation center of the table,
A predetermined illumination region of the surface to be inspected, which is the outer wall surface of the ceramic body, is irradiated with first illumination light and second illumination light having different wavelength bands in one direction perpendicular to the central axis. In a state where the irradiation is performed simultaneously in a mode included in the horizontal plane, the irradiation target area is imaged by the predetermined one imaging unit arranged in the one horizontal plane, and one imaging data of a predetermined data format is converted into one data. An imaging step of generating as an imaging result;
Based on the imaging result in the imaging step, a determination image generation step of generating a determination image that can be used to determine the presence or absence of a crack,
A determining step of determining the presence or absence of a crack on the surface to be inspected based on the image for determination;
With
The first and second illumination lights are arranged so that the angle between the center of each optical axis and the normal to the outer wall has the same angle θ in the range of 5 ° ≦ θ ≦ 30 °. Irradiated on the surface to be inspected from different directions sandwiched,
In the imaging step, while the ceramic body is rotated once around the central axis by rotating the table, imaging is performed by the imaging unit using a surface parallel to the central axis as the surface to be inspected. Thereby, the one imaging data is obtained,
The determination image generation step obtains a pixel value of a predetermined color component from the one piece of image data as an image formation signal for the color component, thereby determining the determination image data that is image data of the determination image. It is a determination image data generation step to be generated,
In the determination image data generating step,
An image consisting of only an image forming signal for the first color component, or an image forming signal for the first color component and an image forming signal for a color component other than the first color component, and First determination image data that is image data of a first determination image that is an image including an image forming signal having a signal amount equal to or less than the threshold value;
An image consisting only of an image forming signal of a second color component having a different wavelength range from the first color component, or an image forming signal of the second color component and a color other than the second color component Second determination image data that is image data of a second determination image that is an image formation signal for the component and is an image formed of an image formation signal having a signal amount equal to or less than a predetermined threshold value;
Is generated,
The first and second determination images are images for which it is possible to determine the presence or absence of a crack on the surface to be inspected based on the difference in the formation of the shadow region in each of the two images when the two are compared,
In the determining step, by comparing the first and second determination image data , it is determined whether or not a shadow area is formed differently in each of the first and second determination images. the determination for the presence of cracks in the inspected surface, is adapted to generate a determination result data that describes the result of the determination,
On the basis of the first and second image data for determination,
In the first and second determination images, when it is determined that there is a shadow region having a smaller pixel value than a periphery extending along the same direction at the same position on the inspection surface of the ceramic body, and,
Although the shadow region exists in one of the first and second determination images, the region corresponding to the shadow region in the other determination image corresponds to any one of the formation position of the shadow region and its vicinity. If it is determined that there is no
It is determined that a crack along the same direction has occurred at a location corresponding to the shadow area on the surface to be inspected of the ceramic body,
A method for inspecting a surface of a ceramic body, comprising: A method for inspecting a surface of a ceramic body according to claim 1,
In the imaging step, an imaging result of the imaging unit is generated as imaging data in a data format in which pixel values of each of a plurality of color components are independently described,
In the determination image data generating step,
By acquiring only the pixel value of the first color component from the imaging data, or by acquiring the pixel value of the first color component and the pixel value of a color component other than the first color component, By obtaining a pixel value equal to or less than a predetermined threshold, the first determination image data is generated,
By acquiring only the pixel value of the second color component from the imaging data, or by acquiring the pixel value of the second color component and the pixel value of a color component other than the second color component, Generating the second determination image data by acquiring a pixel value equal to or less than a predetermined threshold value;
A method for inspecting a surface of a ceramic body, comprising: A method for inspecting a surface of a ceramic body according to claim 1,
In the imaging step, an imaging result by the imaging unit is generated as imaging data in a data format in which pixel value information on a plurality of color components is combined,
In the determination image data generating step,
By decomposing the imaging data, only the pixel value of the first color component is described, or the pixel value of the first color component and the pixel value of a color component other than the first color component are described. Generating the first determination image data in which a pixel value equal to or less than a predetermined threshold value is described,
By decomposing the imaging data, only the pixel value of the second color component is described, or the pixel value of the second color component and the pixel value of a color component other than the second color component are described. Generating the second determination image data, wherein the pixel value is equal to or less than a predetermined threshold value.
A method for inspecting a surface of a ceramic body, comprising: A method for inspecting a surface of a ceramic body according to any one of claims 1 to 3 , wherein
The wavelength range of the first color component at least overlaps with the wavelength band of the first illumination light, and the wavelength range of the second color component at least overlaps with the wavelength band of the second illumination light,
A method for inspecting a surface of a ceramic body, comprising: A method for inspecting a surface of a ceramic body according to any one of claims 1 to 4 ,
In the imaging step, the imaging unit is repeatedly imaged by the imaging unit such that the imaging ranges are adjacent to each other or partially overlap with each other with a predetermined imaging width. Image the entire inspection surface,
Obtaining the first and second imaging results by synthesizing a plurality of captured images obtained by repeated imaging by the imaging unit;
A method for inspecting a surface of a ceramic body, comprising: It is a surface inspection method of the ceramic body of Claim 5 , Comprising:
A line sensor having sensitivity to at least the first and second illumination lights is used as the imaging unit.
A method for inspecting a surface of a ceramic body, comprising: A method for inspecting a surface of a ceramic body according to any one of claims 1 to 6 ,
An image display step of displaying the first and second determination images so that the presence or absence of the crack can be determined by a predetermined image display means;
A surface inspection method for a ceramic body, further comprising: It is a surface inspection method of the ceramic body of Claim 7 , Comprising:
In the determining step, the first and second determination images displayed on the image display unit are compared to form a shadow area in each of the first and second determination images. Determining the presence or absence of cracks in the surface to be inspected by determining whether the surface is inspected. It is a surface inspection method of the ceramic body of Claim 8 , Comprising:
In the determining step, based on the first and second determination images,
When it is determined in the first and second determination images that a shadow area darker than the periphery extending along the same direction exists at the same position on the surface to be inspected of the ceramic body, and
Although the shadow region exists in one of the first and second determination images, the region corresponding to the shadow region in the other determination image corresponds to any one of the formation position of the shadow region and its vicinity. If it is determined that there is no
It is determined that a crack along the same direction has occurred at a location corresponding to the shadow area on the surface to be inspected of the ceramic body,
A method for inspecting a surface of a ceramic body, comprising: A method for inspecting a surface of a ceramic body according to any one of claims 1 to 9 , wherein
A wavelength band of the first illumination light is 400 nm to 500 nm;
A wavelength band of the second illumination light is 600 nm to 800 nm;
A method for inspecting a surface of a ceramic body, comprising: A method for inspecting a surface of a ceramic body according to any one of claims 1 to 9 , wherein
A wavelength band of the first illumination light is 100 nm to 400 nm;
A wavelength band of the second illumination light is 300 nm to 800 nm;
A method for inspecting a surface of a ceramic body, comprising: A method for inspecting the surface of a ceramic body according to any one of claims 1 to 11 ,
The ceramic body is a honeycomb structure obtained by firing a ceramic formed body obtained by extrusion molding, and a side surface of the honeycomb structure is the inspection surface.
A method for inspecting a surface of a ceramic body, comprising: A surface inspection method for a ceramic body according to any one of claims 1 to 12 , wherein:
The imaging step and the determination image generation step,
Said table;
First illuminating means capable of irradiating the first illuminating light to the predetermined illuminated area of the ceramic body held by the holding unit;
A second illumination unit that can irradiate the second illumination light to the predetermined irradiation area of the ceramic body held by the holding unit;
Said imaging means;
A determination image generating unit configured to generate the first determination image data and the second determination image data based on the one imaging result;
Wherein the first and second illumination means are arranged on the one horizontal plane with the imaging means interposed therebetween, and the first and second illumination means are arranged on the surface to be inspected from different directions. A surface inspection method for a ceramic body, which is performed by irradiating light with a surface inspection device.